IT-Integration

Integration Strategies Report

15.05.2002
Zielvorstellung der Integration ist eine flexible IT-Achitektur, die rasch auf geschäftliche Entwicklungen reagieren kann. Dafür ist eine Strategie unabdingbar, argumentieren die Analysten der Butler Group, die sich nicht auf die technische Herausforderung beschränkt, sondern auch Geschäftsprozesse berücksichtigt.

Butler Group believes that integration has become the number one priority for organisations as they strive to create a flexible IT architecture. However, despite the arrival of new technologies, and the assertions of vendors that their products can meet this challenge, research indicates that integration projects are substantially failing. To compound the problem, new initiatives such as CRM and Enterprise Portals, have imposed an additional burden, and organisations are struggling to move beyond an unmanageable jumble of quick-fix solutions.

Business Issues

Integration has to be viewed as a business problem that has both a technical and a business solution. One of the major problems with current integration practice is that it tends to deal with the technical to the exclusion of the business. Only by aligning these two aspects can an organisation build an extensible integration framework, and central to this imperative is an in-depth understanding of the processes that drive the business.

Butler Group believes that integration is running into problems because:

If this predicament is to be overcome, then there must be fundamental changes to integration strategy, and vendors must bridge the wide gap that exists between their product proposition and the reality of where customers are today.

For many enterprises, deploying pragmatic solutions to business integration requirements is a necessity. There are always competitive pressures, the need to bring solutions to market more quickly, that drive short-term solutions. Without a long-term integration strategy, however, companies end up with a series of integration projects that have no overall direction and can have conflicting objectives.

Overall integration costs are substantially higher, because there is little opportunity for reuse, maintenance of the individual solutions is more complex and time-consuming, and each project suffers a recurring overhead of start-up and management expenses. In the worst case, where there have been conflicting objectives for individual projects, it can lead to a tangle of integration links that jeopardise the future development of the IT architecture.

Technology Issues

Three factors are shaping the development of the integration market - business processes, application servers, and packaged integration solutions. Butler Group has, for some time, advocated a business process-driven approach to integration, and the market is now aligned with this view. It is considered essential for integration servers to include a business process layer, and vendors who did not yet have this functionality have obtained it through acquisition or partnership. Companies that have reacted more slowly to this need, including Mercator and iPlanet have lost some ground. Conversely, companies such as SeeBeyond and Vitria whose products have been built from a process perspective have been able to take advantage of this trend.

Another noticeable trend has been the predominance of the application server, and for many vendors, the assimilation of integration technology into a wider infrastructure platform. This applies to IBM, which now brands much of its integration technology (including MQSeries) under the WebSphere banner, and also to BEA, Sybase, and iPlanet amongst others. The J2EE standard supported by these vendors incorporates many of the technologies required for integration, including transaction management, distributed objects, messaging, and XML handling. It also defines the J2EE Connector Architecture (JCA) as a specific interface for application integration.

Several companies, including TIBCO and Vitria, have introduced packaged application suites, which are targeted at departmental projects, and often include pre-packaged templates or patterns for specific business problems. Butler Group believes that these solutions are only to be recommended where they are seen as building blocks in a wider strategy. There is a temptation for customers to believe that they offer a short cut to integration, but whilst they can speed deployment for individual projects, which is to be welcomed, they are not a panacea for all integration requirements.

Integration Strategy

Butler Group believes that the following steps are the key to more effective integration:

Market Analysis

There has been considerable consolidation in the market over the past eighteen months, with a number of high profile acquisitions taking place. We believe that this partly reflects general market conditions, but is also symptomatic of the integration problems that customers have faced. At the time when the dot.com boom still flourished, and public net markets were considered the next big thing, many of the smaller integration vendors were focused on a predominantly B2B integration strategy. Now that this concept has deflated, the centre of attention has moved back to internal application integration, and whilst some of these vendors have made a successful transition, others have been swallowed up by the larger infrastructure vendors.

Butler Group believes that demand in the integration market is exceptionally strong, and will remain so for the next two years, as companies concentrate on integration driven by CRM, Enterprise Portals, and Web-based applications. Beyond that point, underlying market growth will continue, but growth rates will depend on how quickly demand for B2B integration regains momentum. Butler Group forecasts that the market for integration products will exceed US$10 billion by 2007 - a Compound Annual Growth Rate (CAGR) of 26.4%.

The leading vendors in the integration market are BEA, GXS, IBM, iPlanet, Microsoft, Mercator, Peregrine, TIBCO, SeeBeyond, Sybase, Vitria, and webMethods. IBM has the largest market share, and its acquisition of CrossWorlds will help it to maintain this position. Butler Group believes that the companies that will outperform the market in 2002 are BEA, Microsoft, TIBCO, and webMethods. There is likely to be further consolidation in the market over the next year, particularly if overall IT spend takes longer than expected to recover.

Conclusions

The goal of end-to-end integration is an agile and flexible IT architecture that can respond quickly to business needs. For most companies this is a distant dream, and integration is seen as a problem to be solved, rather than an opportunity to add value to the business. The key to changing this is to adopt a strategic approach to integration, that provides the technical and business framework to guide and measure individual projects.